diff --git a/proxmox/cluster/README.md b/proxmox/cluster/README.md
index a07c098..5224767 100644
--- a/proxmox/cluster/README.md
+++ b/proxmox/cluster/README.md
@@ -352,23 +352,47 @@ Pentru detalii despre:
 
 ## Backup Job Configuration
 
-### Job ID: backup-fbb668c0-726e
-- **Schedule:** Daily la 02:00
-- **Compression:** zstd
-- **Mode:** snapshot
-- **Storage:** backup
-- **VM-uri incluse:** 100, 103, 104, 106, 108, 201
-- **Retention:** 1 daily, 1 weekly
-- **Fleecing:** Disabled
-- **Notes Template:** {{guestname}}
+Nu există un job unic, ci **8 job-uri separate**, alternate pe zile pare/impare ca să nu se
+suprapună sarcina termică și de I/O. Sursa de adevăr e `/etc/pve/jobs.cfg`
+(`pvesh get /cluster/backup`), nu tabelul de mai jos — verifică-l dacă ceva nu se potrivește.
 
-**Containere în backup:**
-- 100 (Portainer) - 10.0.20.170
-- 103 (Dokploy) - 10.0.20.167
-- 104 (Flowise) - 10.0.20.161
-- 106 (Gitea) - 10.0.20.165
-- 108 (Oracle) - 10.0.20.121
-- 201 (Windows 11 VM) - DHCP
+| Job | Guest | Când | Storage | Retenție |
+|---|---|---|---|---|
+| `backup-impare-oracle` | 108 (Oracle) | zile impare, 02:00 | `backup` | keep-last=2, keep-weekly=1 |
+| `backup-impare-rest` | 100, 106, 171 | zile impare, 03:00 | `backup` | keep-last=2, keep-weekly=1 |
+| `backup-impare-adina` | 303 (Win11-Adina) | zile impare, 03:30 | `backup` | keep-last=2, keep-weekly=1 |
+| `backup-pve1-minecraft` | 101 | zile impare, 22:00 | `backup-ssd` | keep-last=2, keep-weekly=1 |
+| `backup-pare-flowise` | 104 | zile pare, 02:00 | `backup` | keep-last=2, keep-weekly=1 |
+| `backup-pare-windows` | 201 (roacentral) | zile pare, 03:30 | `backup` | keep-last=2, keep-weekly=1 |
+| `backup-pve1-moltbot` | 110 | zile pare, 22:00 | `backup-ssd` | keep-last=2, keep-weekly=1 |
+| `backup-f1c87584-8eb2` | 109 (DR) | prima sâmbătă din lună, 15:00 | `backup-pvemini-nfs` | keep-monthly=1 |
+
+Toate rulează cu `mode snapshot`, `compress zstd`, `notes-template {{guestname}}`.
+Job-urile sunt executate de daemonul `pvescheduler` (nu de un timer systemd — nu căuta
+`pvescheduler.timer`, nu există):
+
+```bash
+systemctl is-active pvescheduler
+pvesh get /cluster/backup          # lista reala a job-urilor
+```
+
+### ⚠️ Guest-uri FĂRĂ backup
+
+| Guest | Stare | Comentariu |
+|---|---|---|
+| **102 `docker.romfast.ro`** | **running** | **neacoperit — Docker host + sandbox-uri agenți AI** |
+| **103 `dokploy`** | **running** | **neacoperit** |
+| 301 `docker-portainer-template` | stopped | template, recreabil |
+| 302 `oracle-test-302` | stopped | mediu de test, recreabil din 310 |
+| 310 `Win11-Template-Sysprep` | stopped | template — de evaluat dacă merită backup |
+
+Cele două `running` sunt singurele care contează. Verificare rapidă a acoperirii:
+
+```bash
+pvesh get /cluster/resources --type vm --output-format json | \
+  python3 -c "import sys,json;[print(r['vmid'],r.get('name'),r['status']) for r in json.load(sys.stdin)]"
+pvesh get /cluster/backup --output-format json | grep -o '\"vmid\":\"[^\"]*\"'
+```
 
 ### Comenzi Cluster